A spatiotemporal data mining study to identify high-risk neighborhoods for out-of-hospital cardiac arrest (OHCA) incidents

Abstract

Out-of-hospital cardiac arrest (OHCA) is a worldwide health problem. The aim of the study is to utilize the territorial-wide OHCA data of Hong Kong in 2012–2015 to examine its spatiotemporal pattern and high-risk neighborhoods. Three techniques for spatiotemporal data mining (SaTScan’s spatial scan statistic, Local Moran’s I, and Getis Ord Gi*) were used to extract high-risk neighborhoods of OHCA occurrence and identify local clusters/hotspots. By capitalizing on the strengths of these methods, the results were then triangulated to reveal “truly” high-risk OHCA clusters. The final clusters for all ages and the elderly 65+ groups exhibited relatively similar patterns. All ages groups were mainly distributed in the urbanized neighborhoods throughout Kowloon. More diverse distribution primarily in less accessible areas was observed among the elderly group. All outcomes were further converted into an index for easy interpretation by the general public. Noticing the spatial mismatches between hospitals and ambulance depots (representing supplies) and high-risk neighborhoods (representing demands), this setback should be addressed along with public education and strategic ambulance deployment plan to shorten response time and improve OHCA survival rate. This study offers policymakers and EMS providers essential spatial evidence to assist with emergency healthcare planning and informed decision-making.